Administrator Guide

Virtual Port Mode
Virtual port mode provides port and controller redundancy by connecting multiple active ports to each Fibre Channel or Ethernet
switch.
In virtual port mode, each physical port has a WWN (World Wide Name), and is also assigned an additional virtual WWN. Servers
target only the virtual WWNs. During normal conditions, all ports process IO. In the event of a port or controller failure, a virtual
WWN will move to another physical WWN in the same fault domain. When the failure is resolved and ports are rebalanced, the
virtual port returns to the preferred physical port.
Virtual port mode provides the following advantages over legacy mode:
• Increased performance: Because all ports are active, additional front-end bandwidth is available without sacrificing redundancy.
• Improved redundancy: Ports can fail over individually instead of by controller.
• Simplified iSCSI configuration: Each fault domain has an iSCSI control port that coordinates discovery of the iSCSI ports in the
domain. When a server targets the iSCSI port IP address, it automatically discovers all ports in the fault domain.
ALUA Port Mode
Asymmetric Logical Unit Access (ALUA) provides port and controller redundancy for SAS front-end connections.
Volumes mapped to a server using SAS front-end also have port and controller redundancy. Volumes mapped over SAS are mapped
to both controllers. The volume mapping is Active/Optimized on one controller and Standby on the other controller. If the port or
controller fails on the active controller, the paths to the other controller become Active/Optimized. The mapping on the first
controller switches to Standby. When the port or controller recovers, the mapping to the first controller returns to Active/
Optimized and the mapping to the second controller returns to Standby status.
Legacy Mode
Legacy mode provides controller redundancy for a dual-controller Storage Center by connecting multiple primary and reserved ports
to each Fibre Channel or Ethernet switch.
In legacy mode, each primary port on a controller is paired with a corresponding reserved port on the other controller. During normal
conditions, the primary ports process IO and the reserved ports are in standby mode. If a controller fails, the primary ports fail over
to the corresponding reserved ports on the other controller. This approach ensures that servers connected to the switch do not
lose connectivity if one of the controllers fails. For optimal performance, the primary ports should be evenly distributed across both
controllers. When possible, front-end connections should be made to separate controller IO cards to improve redundancy.
Fault Domains
Front-end ports are categorized into fault domains that identify allowed port movement when a controller reboots or a port fails.
Failure modes and port activity depend on whether the Storage Center is configured for Legacy mode, ALUA port mode, or Virtual
port mode.
Fault Domains for SCv2000 Series Controllers
When used on SCv2000 series controllers, Storage Center handles all fault domain creation and modification.
Fault domain behavior on SCv2000 series controllers:
• Fault domains are automatically generated.
• There are always two fault domains for IO in Fibre Channel and iSCSI configurations, not including replication-only domains.
• Fault domains are automatically created for Flex/Embedded Ethernet ports.
• Four fault domains are created for front-end SAS ports.
NOTE: Fault domains cannot be modified by users with SCv2000 series controllers.
Storage Center Maintenance
93